Abstract
Data pre-processing plays an important role in data mining for ensuring good quality of data especially dealing with industrial datasets. This work presents an exemplar case study for the prediction of the inclusions population in steel products, which demonstrates the importance of variable selection to obtain satisfactory classification accuracy and to achieve a deep understanding of the phenomenon under consideration. A novel variable selection approach has been applied for selecting the variables which mainly affect the target, preliminary to the design of the classifier. Five different classifiers have been designed and applied and the obtained results are presented, compared and discussed.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Borselli, A., Colla, V., Vannucci, M., Veroli, M.: An image inference system applied to defect detection in flat steel production. IEEE World Congr. Comput. Intell. 1, 1–6 (2010)
Cateni, S., Colla, V., Vannucci, M.: General purpose input vvariable extraction: a genetic algorithm based procedure give a gap. In: Proceedings of the 9th International Conference on Intelligence Systems design and Applications ISDA’09, pp. 1278–1283 (2009)
Cateni, S., Colla, V., Vannucci, M.: Variable selection through genetic algorithms for classification purpose. In: Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010, pp. 6–11 (2010)
Cateni, S., Colla, V., Vannucci, M.: A genetic algorithm-based approach for selecting input variables and setting relevant network parameters of som-based classifier. Int. J. Simul. Syst. Sci. Technol. 12(2), 30–37 (2011)
Cateni, S., Colla, V., Nastasi, G.: A multivariate fuzzy system applied for outliers detection. J. Intell. Fuzzy Syst. 24(4), 889–903 (2013)
Cateni, S., Colla, V., Vannucci, M.: A hybrid feature selection method for classification purposes. In: Proceedings on UKSim-AMSS 8th European Modelling Symposium on Computer Modelling and Simulation, EMS 2014, At Pisa (Italy), pp. 39–44 (2014)
Duda, R., Hart, P.: Pattern Classification and Scene Analysis. Wiley, New york (1973)
ECSC supported project: Quality improvement for continuously cast steel products by swirling flow strategies. Technical report, Contract number 7210-PR (211)
Fageth, R., Allen, W., Jager, U.: Fuzzy logic classification in image processing. Fyuzzy Syst. 82(3), 265–278 (1996)
Fausett, L.: Foundaments of Neural Networks. Prentice Hall, New York (1994)
Haykin, S.: Self Organizing Maps in Neural Networks—A Comprehensive Foundation, 2nd edn. Prentice Hall (1999)
Komperod, M., Hauge, T., Lie, B.: Preprocessing of experimental data for use in model building and model validation. In: The 49th Scandinavian Conference on Simulation and Modeling (2008)
Kwon, Y.J., Zhang, J., Lee, H.G.: A cfd based nucleation-growth-removal model for inclusion behaviour in a gas agitated ladle during molten steel deoxidation. ISIJ Int. 38, 6 (2008)
Lopes, N., Ribeiro, B.: A data pre-processing tool for neural networks (dptnn) use in a moulding injection machine. In: Proceedings of the Second World Manufacturing Congress Durham, pp. 357–361. England (1999)
Sheng, D.Y., Soder, M., Alexis: Most relevant mechanism of inclusion growth in an induction-stirred ladle. Scand. J. Matall. 31, 210–220 (2002)
Sheng, D., Soder, M., Jönsson, P.: J. Scand. J. Metall. 31, 134–147 (2002)
Sokolova, M., Lapalme, G.: A systsystem analysis of performance measures for classification tasks. Inf. Process. Manage. 4, 427–437 (2009)
Taguchi, K.: Complex deoxidation equilibria in molten iron by aluminium and calcium. ISIJ Int. 45, 1572–1576 (2005)
Vapnik, V.: Statistical Learning Theory. Wiley, New York (1998)
Zhang, L., Thomas, B.G., Wang, X., Cai, K.: Evaluation and control of steel cleanliness review. In: 85th Steelmaking Conference Warrendale, PA, pp. 431–452 (2002)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this paper
Cite this paper
Cateni, S., Colla, V. (2016). The Importance of Variable Selection for Neural Networks-Based Classification in an Industrial Context. In: Bassis, S., Esposito, A., Morabito, F., Pasero, E. (eds) Advances in Neural Networks. WIRN 2015. Smart Innovation, Systems and Technologies, vol 54. Springer, Cham. https://doi.org/10.1007/978-3-319-33747-0_36
Download citation
DOI: https://doi.org/10.1007/978-3-319-33747-0_36
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-33746-3
Online ISBN: 978-3-319-33747-0
eBook Packages: EngineeringEngineering (R0)